Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement

By Risheng Liu, Long Ma, Jiaao Zhang, Xin Fan, Zhongxuan Luo. Computer Vision and Pattern Recognition, 2020

1,024 citations recorded by the source corpus.

Low-light image enhancement plays very important roles in low-level vision areas. Recent works have built a great deal of deep learning models to address this task. However, these approaches mostly rely on significant architecture engineering and suffer from high computational burden. In this paper, we propose a new method, named Retinex-inspired Unrolling with Architecture Search (RUAS), to construct lightweight yet effective enhancement network for low-light images in real-world scenario. Specifically, building upon Retinex rule, RUAS first establishes models to characterize the intrinsic underexposed structure of low-light images and unroll their optimization processes to construct our holistic propagation structure. Then by designing a cooperative reference-free learning strategy to discover low-light prior architectures from a compact search space, RUAS is able to obtain a top-performing image enhancement network, which is with fast speed and requires few computational resources. Extensive experiments verify the superiority of our RUAS framework against recently proposed state-of-the-art methods. The project page is available at http://dutmedia.org/RUAS/.

Subfields: Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/4306e545978bb653b2c9ac1a6f704428482877b6

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Low-Light Image and Video Enhancement Using Deep Learning: A Survey (2021, methodology): “Zero-shot learning-based methods : ExCNet [27], Zero-DCE [28], RRDNet [29], Zero-DCE++ [30], RetinexDIP [31], and RUAS [32].”
Retinexformer: One-stage Retinex-based Transformer for Low-light Image Enhancement (2023, methodology): “Plain methods like histogram equalization and gamma correction tend to produce undesired artifacts because they barely consider the [49]), ReNet (RetinexNet [54]), KinD [66], and RUAS [30] on six low-light image enhancement benchmarks. illumination factors.”
Toward Fast, Flexible, and Robust Low-Light Image Enhancement (2022, neutral): “SCI can be applied to improve the performance for existing works, e.g., RUAS [14].”
R2RNet: Low-light Image Enhancement via Real-low to Real-normal Network (2021, methodology): “[18] proposed a Retinex-inspired Unrolling with Architecture Search (RUAS) and designed a cooperative reference-free learning strategy to discover low-light prior architectures from a compact search space.”
Target-aware Dual Adversarial Learning and a Multi-scenario Multi-Modality Benchmark to Fuse Infrared and Visible for Object Detection (2022, background): “Deep learning has achieved promising progress in low-level vision tasks [12, 15, 17, 19, 23, 25, 35, 40] due to the powerful nonlinear ﬁtting ability of multi-layer neural networks.”
SNR-Aware Low-light Image Enhancement (2022, methodology): “612 Methods LPNet [26] MIR-Net [59] RF [24] 3DLUT [60] A3DLUT [42] Band [52] EG [20] Retinex [29] Sparse [53] IPT [6] Uformer [44] Ours PSNR 19.”
URetinex-Net: Retinex-based Deep Unfolding Network for Low-light Image Enhancement (2022, methodology): “Furthermore, to verify the efficiency of our model in learning implicit prior from data, we compare with state-ofthe-art (SOTA) learning-based methods including RetinexNet [34], KinD [41], Zero-DCE [11], RUAS [24], AGLLNet [26] and KinD++ [40].”
Ultra-High-Definition Low-Light Image Enhancement: A Benchmark and Transformer-Based Method (2022, methodology): “…et al. 2018), DSLR (Lim and Kim 2020), KinD (Zhang, Zhang, and Guo 2019)), two unsupervised CNN-based methods (EL-GAN (Jiang et al. 2021), RUAS (Liu et al. 2021b)), two zero-shot learning-based methods (Z_DCE (Guo et al. 2020), Z_DCE++ (Li, Guo, and Loy 2021) ) and two supervised…”
Learning a Simple Low-Light Image Enhancer from Paired Low-Light Instances (2023, methodology): “7, the reflectance contains rich textures and details while the illumination is piece-wise continuous and texture-less, demonstrating PairLIE can accurately decompose the [9], RetinexNet [32], GALDNet [30], KinD [38], URetinexNet [33], Zero-DCE [7], RRDNet [41], RUAS [18], SCI [22], EnlightenGAN [12”
Image De-Raining Transformer (2022, neutral): “…tasks [17], [18], [31], [32], [33], [34], [35], such as de-noising [36], [37], [38], super-resolution [39], [40], [41], de-blurring [42], [43], [44], [45], de-hazing [19], [19], [46], [47], [48], [49], [50], low-light image enhancement [51], [52], [53], [54], [55], style transfer [56], [57], etc .”